Dell and Nvidia have announced they are teaming up in a bid to help deliver AI-powered hardware solutions for enterprises.
The two tech giants are merging their areas of expertise into a single package, allowing businesses to benefit from the power of generative AI – including LLMs like the popular AI writer ChatGPT – on premises.
Dubbed Project Helix, the initiative will also provide validated blueprints and reference deployments to help firms get their AI workloads off the ground.
Looking at the hardware, Dell will be providing its PowerEdge servers, such as the XE9680 and R760a, and Nvidia will be providing its new H100 tensor core GPUs, which succeeds the A100 used to power ChatGPT.
The hardware will integrate with Dell PowerScale and Dell ECS, the company’s storage solutions, and Nvidia will also provide its Nvidia AI Enterprise software, as well as making use of the Nvidia NeMo framework for generative AI.
Kari Briski, VP of software product management at Nvidia told VentureBeat that with this on-premise solution to AI workloads, “we’re bringing security and privacy to enterprise customers.” Most commonly, enterprise will be using AI that is hosted in the cloud, not on premises.
She added, “Every enterprise needs an LLM for their business, so it just makes sense to do this on premises.”
Even though Briski conceded that generative AI may be overhyped in some cases, she wanted the new joint venture between the two companies to enable what she referred to as LLMops – making LLMs operational for practical enterprise purposes.
“What we haven’t been doing is providing these pre-trained foundational models in a way that’s easily replicable,” she said.
The blueprints for Project Helix will guide enterprise on customizing their AI deployment to meet their specific needs. Varun Chhabra, SVP for product marketing at Dell thinks it is critical to get the right mix of computer resources, and that Project Helix provides the best ways to achieve this.
In addition to on premises, Chhabra also believes that with Dell hardware, enterprise can benefit from AI deployed in the cloud as well.
Designs based on Project Helix will be available from July this year, via traditional channels and APEX flexible consumption options.